Computational Models of Its Learning Circuit

The learning phenomenon allows various analysis levels, but this article treats one specific paradigm of Artificial Intelligence, Artificial Neural Networks, whose main virtue is their capacity to look for unified and mutually satisfactory solutions with an important turn towards biological and psychological models. Given the fact that a substantial part of the procedures, methods, etc. proposed until the present time use the principles, models and data of biology and psychology, or both at the same time, we focus on models which look for a greater degree of coherence. That is the reason why this article analyzes and compares all the aspects comprised by two Artificial Neural Networks models whose implementation is presented: Gluck’s and Thompson’s, and Hawkins’. A multithread computer model is developed with the purpose of analyzing those models in order to study the simple learning phenomena in a sea invertebrate, the Aplysia Californica, and check their capacity for research in psychology and neurobiology. The predictive capacity differs significantly for both models.: the Hawkins model covers better the behavioral repertory of Aplysia, on the associative as well as the non-associative learning level. Through the integration with neurobiological and behavioral models of associative learning, the applied Artificial Neural Networks modelling technique broadens its scope, allowing the enhancement of some architectures and procedures that are being used nowadays.